1242 
still elements that ave inclined with respect to the axis of the tube, 
but the inclination has decreased considerably. 
4. Now we pass to a second experiment. A strong light beam, 
which over a length of a few centimeters has a diameter of 4mm., 
ean easily be thrown along the axis of the tube “D. The velocity 
of the small gas-bubbles illuminated by this light-beam was 
measured by means of a rotating mirror the axis of which was 
parallel to that of the glass tube. 
If the mirror is at rest we see in a telescope, pointing to the 
mirror, the streamlines principally in horizontal direction. From the 
inclination which the apparent stream-lines assume when the mirror 
rotates with a known velocity we can derive the value of the 
velocity in the axis of the tube i.e. the maximal velocity. It is evident, 
that this determination is most accurate when the velocity of the 
mirror and the distance from the glass tube are so chosen that 
the angle of inclination « of the stream-lines with the horizontal 
direction becomes nearly 45°. By reversing the current in the electro- 
motor, by which the mirror is rotated we can directly read the angle 
2a. The velocity of the mirror was 1,052 rotations per second. The 
“effective length” of the distance of the axis of rotation of the mirror 
to the axis of the tube could be determined within a fraction of a 
millimeter, attention being paid of course to the passage of the 
light through water and through glass. 
5. In a direction perpendicular to the tube the small gasbubbles 
scatter less light than in a direction which makes a smaller angle 
with the water-current. 
In some experiments I have observed in an ‘clined direction, the 
greater intensity in which is advantageous for the accuracy of the adjust- 
ments. But then the observations must be reduced to the values that 
would be found in a plane perpendicular to the axis of rotation of 
the mirror. Instead of calculating this reduction from the angles that 
determine the deviation I have preferred to determine experimentally 
by a separate experiment the corrections for the observed values of «. 
For this purpose | used a series of luminous points, which were 
moving with the velocity of the water-current, but exhibiting 
a greater intensity and a particular regularity. Thirty-six small steel 
balls were fixed near the circumference of a copper disk of about 
20 em. diameter. In this additional experiment the plane of the disk 
could be placed perpendicularly to the direction in which in the 
flow-experiments the light from the tube fell on the mirror. At 
